A new method for high throughput screening has been developed based upon the inclusion of the target receptor, chaperon protein, or nuclear receptor in a flowing chromatographic system. In this approach, the target protein is immobilized on a solid support and the support packed into a small column. The test chemicals are passed through the column, over the immobilized target, and the time that it takes for the compounds to pass from the beginning of the column to its end is directly related to the strength of interaction between the target and the compound, i.e. the binding affinity of the ligand-receptor complex. Using this method, complex chemical and biological mixtures can be rapidly sorted between compounds that interact and do not interact with the disease-related target. At the same time, the compounds that bind to the target are themselves rapidly sorted between low, medium and high affinity binders. Thus, the method quickly provides a large amount of data with high information content. We have developed columns containing G-protein coupled receptors (GPCR) including the kappa, mu and delta subtypes of the opioid recpetor, beta2-adrenergic receptors, various subtypes of the nicotinic receptor and columns containing multiple GPCRs including a histamine-1 and P2Y1 receptor column and a multiple ligand-gated ion channel column containing the alpha 7 and alphaXbetaY nicotinic acetylcholine receptors, NMDA and GABA(A)receptors. The nicotinic receptor-based columns have been used to screen tobacco smoke condensates and initial results indicate that previously unknown compounds have been identified. These compounds are being assessed for their pharmacological activity as competitive agonists and antagonists and non-competitive inhibitors. A study using non-competitive inhibitors and non-linear chromatography was conducted and demonstrated that the method can be used to identify and characterize the non-competitive inhibitors which bind in the central lumen of the receptor as well as at an extra-receptor site identified as the quinacrine binding site. Chemometric analysis was used to develop quantitative structure-activity relationships (QSAR) and the resulting equations can be used to predict the pharmacological activity of a compound. Molecular modeling studies were used to describe and predict the interactions between the nicotinic receptors and non-competitive inhibitors. The computer-based studies have been used as the basis for a patent application entitled Computer-based model for identification and characterization for non-competitive inhibitors of nicotinic acetylcholine receptors and related ligand-gated ion channel receptors (ref number: E-158-2003/0-US-01). The general approach has been expanded through the immobilization of heat shock 90 (Hsp90) protein in the surface of magnetic beads. The Hsp90-beads were used to extract small molecules that bind to Hsp90 from chemical and botanical mixtures and to extract proteins that form complexes with Hsp90 from protein mixtures. The beads were also used to extract proteins that bind to Hsp90 from cellular extracts and indicate that the approach can be used to study intra-cellular protein-protein interactions and to identify new pathways associated with Hsp90. We have recently demonstrated that astrocytoma and glioblastoma call lines express the ERRa and ERRg and that these nuclear receptors are key targets in the treatment of brain cancer. The LC-ERRa and LC-ERRg columns were created and characterized and will be used to screen chemical mixtures and biological matrices for new lead drug candidates.